Reaction kinetics & Chemical Reaction Models Ivan A. Gargurevich, Ph.D. ?

2001
Reproduced at The Chemical Engineers Resource Page with the express permission of the author. The author retains all copyrights to the original work. The version of this article formatted for the web is copyrighted by The Chemical Engineers Resource Page (www.cheresources.com), 2001

Reaction Kinetics & Chemical Reaction Models The fundamentals of chemical reaction kinetics will be presented with the purpose of building, starting with elementary reactions, complex mechanisms. These mechanisms consisting of many elementary reactions can be used with existing software (discussed later) to make predictions on the performance of chemical reactors with special consideration to the formation of trace species. The focus will be on homogeneous processes taking place in the gas phase. The majority of the material presented in this manuscript is based on the authors own research (Gargurevich, 1997). Both in past and present literature dealing with the design of chemical reactors, there is an oversimplification of the chemical reaction models, with the use of global mechanisms consisting of a few reactions with empirically determined reaction rates (Worstell, 2001; Arakawa et al., 1998). For example, the rate of consumption of reactant A by B to form product C, represented by the overall reaction (1) below, is presented in the form of equation (2), A +B = => C
a b d ?A? ? A exp ? E e dt RT ?A? ?B ?

(1)

(2)

where A e is the pre-exponential factor, E an empirically determined activation energy, and a and b, the exponents of the reactant concentrations that are able to represent the concentration dependence over a range of conditions also empirically determined. Unfortunately, these oversimplified mechanisms, to give an example, may not be able to accurately predict the formation of toxic products present in very small concentrations because depending on conditions their formation is dependent on complex chemistry involving stable and radical species, as well as reaction temperature. There is a need then to arrive at more complex mechanisms consisting of elementary reactions that are relevant to the consumption of the reactants, formation of intermediate species and products, and any other chemical species of interest (Senkan, 1992).

General Features of Reaction Mechanisms There are fundamental features in most reaction mechanisms occurring in the gas phase that can be used to assemble such mechanisms. The most important one is the role played by free radicals in the decomposition of any reactant. A very simple example found in freshman chemistry books is the formation of hydrogen iodide starting from hydrogen and iodine (Chang, 1994). The proposed mechanism is a two-step mechanism, I2 < = = > 2I H2 + 2I < = = > 2HI (3) (4)

The first step involves the formation of iodine radicals in reaction (3), and it is followed by the formation of hydrogen iodide. Another example is the oxidation of hydrocarbons at low temperatures initiated by the following reactions (Dryer, 1991) RH + O2 < = = > R +HO2 +M R + O2 <= = > RO2 (5) (6)

in which the hydrocarbon radical R plays a role in reaction (6) forming a species that can lead to the formation of oxygenated species such as alcohols and aldehydes. These types of reaction seem to take place in the urban atmosphere (Seinfield, 1989). If the temperature is high enough, the decomposition of the reactant is initiated by a thermal decomposition such as reaction (7), CH4 <= = => CH3 + H (7)

Reaction (7) is endothermic and its extent would depend on temperature, and results in the formation of methyl radical. This type of reaction would occur in shock tubes, for example, where the decomposition of the fuel is initiated by a traveling shock wave causing temperature and pressure to rise considerably in a very short time. The above examples then illustrate a very important point to remember when examining reaction mechanisms: the decomposition of the reactant is initiated by formation of radical species that can later participate in the reactions leading to the product. The concentration of the radicals formed increase as the reaction time or temperature increase.

Another important consideration is the formation of chain reactions. The basic premise of chain reaction mechanisms is also that free radicals play a leading role in the destruction of reactant molecules. The chain reaction mechanism itself consists of several steps: initiation, propagation, branching (not always present), and termination. This can be illustrated, for certain range of temperature and pressure, by some of the reactions in the following Hydrogen oxidation mechanism: Initiation Propagation Branching H2 + O2 < = = > 2OH H2 + OH < = = > H2O + H H + O2 < = = > OH + O H2 + O < = = > H + OH H2O + O < = = > OH + OH + OH + M < = = > H2O + M Plus Others The initiation step is responsible for the initial decomposition of the reactants. Propagation steps involve a radical-molecule reaction with generation of a radical in the product, while branching reactions produce two radicals per single radical in the reactants. For example, the generation of a flame in combustion is due to branching reactions predominating over termination, with a large generation of radicals resulting in the fast decomposition of the fuel. To illustrate what has been discussed above, in assembling the main reaction paths for the combustion of methane ( Miller et al., 1992) consideration is given to the following reactions: CH4 + X < == > HX + CH3 CH3 + O2 < = = > CH2O + H CH2O + X < = = > CHO + HX HCO + M < = = > H + CO + M HCO + O2 < = = > CO + HO2 HCO + OH < = = > CO + H2O HCO + H < = = > CO + H2 CO + OH < = = > CO2 + H (14) (15) (16) (17) (18) (19) (20) (21) (8) (9) (10) (11) (12) (13)

Termination

Where X represents the free radicals H, OH, or O. The relative concentration of the radicals depend on the stoichiometry of combustion for example, in fuel rich flames, the chemistry of H radicals gains much importa nce. To the above reactions one must add the very important chain branching reaction,

H + O2 < = = > OH + O

(10)

In examining this simplified mechanism, the main conclusions are that at the high temperature of combustion the decomposition of methane is initiated by radical attack such as in reaction (14) since the concentration of free radicals is relatively abundant. Reactions (15), (16), (18), (21) propagate the decomposition of methane. Reactions (17), (10) are chain branching reactions. Reactions (19), (20) lead to stable molecules and would be considered as termination reactions. Obviously, the radical pool is initiated by the thermal decomposition of methane in the hot region of the flame or reaction (22) below, CH4 + M < = = > CH3 +H + M (22)

with the very important branching reaction (10) adding to the radical pool. To summarize, reaction mechanisms can be assembled from elementary reactions using free radicals as the means for decomposition of the reactant, and intermediate products. Chain branching reactions, if they occur, take a very important role in the mechanism as they lead to the formation of increasing concentrations of radicals. Reaction time and temperature have a bearing on radical concentration, and the type of reaction i nitiating the consumption of the reactant. Estimation of Kinetic Coefficients Elementary chemical reactions can be classified as either energy-transfer limited or chemical reaction rate limited (Senkan, 1992). In energy transfer-limited processes, the observed rate of reaction corresponds to the energy transfer to or from species either by intermolecular collisions or by radiation, or intramolecularly due to energy transfer between different degrees of freedom of a chemical species. All thermally activated unimolecular reactions become energytransfer limited at low-density conditions because the reactant can receive the necessary activation energy only by intermolecular collisions. The reaction then becomes pressure dependent at a given temperature. An example of such reaction is the thermal decomposition of hydrogen, H2 + M < = = > H + H + M (23)

The energy for the reaction to occur is generated by collision with a second body M.

Chemical rate limited processes, in the other hand, correspond to chemical reactions occurring under conditions in which the statistical distribution of molecular energies obey the Maxwell-Boltzman form, i.e., the fraction of molecules that have an energy E or larger is proportional to e - E/RT . The rates of intermolecular collisions are very rapid and all species are in equilibrium with the gas mixture. Table 1 depicts several theories that can be applied to estimate rate coefficients in order of increasing complexity. In the simplest approach, the rate coefficient of a bimolecular reaction is simple the collision frequency between the molecules. To improve upon this approximation, the collision frequency needs to be corrected to account for the fact that only those collisions with energies above the activation energy of the reaction will result in a net reaction. Also, a steric factor has to be included, since only collisions taking place in a given spatial arrangement will lead to a net reaction. The next level of complexity is Transition State Theory (TST) of both unimolecular and bimolecular reactions. In TST, the rate coefficients include an activation energy factor, and an entropy factor to account for steric factors. TST only applies to chemical rate limited processes. The Lindemann approach to unimolecular reactions would fall within this level of complexity. Finally, the most complex theories involve the quantum mechanical treatment of energy transfer limited processes such as thermal activation and unimolecular/bimolecular chemical activation. By chemical activation, in the case of a bimolecular reaction for example, is meant that as the result of a bimolecular reaction an intermediate species is formed possessing excess energy over the ground state that can more easily lead to some final product by decomposition, A + B < = = > Activated Molecule < = = > Products These quantum theories account for the dependence of the overall rate coefficients on the excess vibrational energy of the molecular species. In reaction modeling rate coefficients are normally expressed in the modified Arrhenius form,
? ? Ea ? n k ? AT exp ? ? ? RT ?

(24)

A is the collision frequency factor, T is the temperature (the exponent n accounts for non-Arrhenius behavior to fit experimental data) and E a is the activation

energy. Non-Arrhenius behavior is most obvious in reactions that have little activation energies with the pre-exponential factor determining the temperature dependence. The discussion above has established an approach for building more complex reaction mechanisms, and the theoretical foundations for the estimation of reaction rate coefficients. The most important consideration always is the chemistry included in the mechanism. Assembling the elementary reactions composing the mechanism is followed then by the best assessment for the mathematical expression giving the rate coefficients of each reaction. The procedure to follow based on the authors experience is discussed below.

1. Literature Data & Order-of-Magnitude Estimates. Very often, the best value to use for the rate coefficient of the reaction is the literature value, i.e. experimentally determined coefficient, if available. Consideration must be given to the temperature and pressure conditions since as the discussion above has illustrated, they have an effect on the rate coefficient. This is true, for example, in the case of unimolecular reactions, and chemically activated reactions. There several sources of chemical kinetic data, some of which are as follows: (a) High Temperature Reactions- (for example, Methane Combustion) in chronological order: Baulch et al. (1992), Wang (1992), Miller & Bowman (1989), Warnatz (1984), Westbrook & Dryer (1984). (b) Chlorinated Hydrocarbons- High Temperature Reactions: Qun & Senkan (1994), Senkan (1993). (c ) For lower temperatures and reactions occurring in the ambient air: DeMore et al. (1990). An extensive data base for chemical reaction kinetics can be found in the National Institute of Standards Chemical Kinetics Database (1994). The reference can be found at the end of the manuscript. Sometimes, a rough-order-of magnitude value for the rate constant is needed for two reasons: either no other value is available or the aim is to scan the mechanism for reactions that have small impact on the consumption of the

reactant, product formation, or formation of any other species of interest. One way to make such an estimate is by the method of analogous reactions as depicted in Table 2 taken form Senkan (1992). On inspection of Table 2, there are several issues that are of importance in estimating rate coefficients. Unimolecular fission reactions are endothermic, and the heat of reaction corresponds to the minimum activation energy that could be expected for the reaction. Simple kinetic theory of bimolecular reactions gives the following expression for the rate coefficient of the reaction between molecules A and B ( Laidler, 1987),

? 8kT ? 2 k ? Z AB ? N A? ? AB ? 2 ? cm3/mol-sec (25) ?? ? ? ? ? ZAB is the molar collision frequency, ? AB is the mean collision diameter or rigid sphere collision cross-section, ? AB is the reduced mass, NA is Avogadros number, and k is Boltzmanns constant. The estimate of the molar collision frequency at 300 oK turns out to be 1.0 x 1013 cm3/mol-sec, and it represents the upper limit for the bimolecular rate coefficient without accounting for the activation energy or steric factor. As stated above, a lower limit for the activation energy for endothermic reactions is the heat of reaction.
2

Another method that can be used to estimate the activation energy of metathesis reactions such as, H + CH4 < = = > CH3 + H2 is the Evans -Polanyi relationship for similar reactions or, E ? ? ?? ? H rn ?? ? (26) where -? Hrn is the heat of reaction, which is defined as positive for an exothermic reaction, and ? and ? are the Evans-Polanyi empirical constants for the family of reactions. Polanyi relationships often fail when there is charge separation involved in the transition state, such is the case when atoms or groups involved in the reaction differ in electronegativities. 2. Transition State Theory of Unimolecular/Bimolecular Reactions. A chemical reaction is presumably a continuous process involving a gradual transition from reactants to products. It has been found extremely helpful, however, to consider the arrangement of atoms at an intermediate stage of reaction as though it were an actual molecule. This intermediate structure is the

transition state, and its energy content corresponds to the top of the reaction energy barrier along the reaction coordinate. The rate coefficients according to TST will be given here without proof, they correspond to a thermodynamic approach where the reaction rate is given in terms of thermodynamic functions. One of the main assumptions of TST is that the process is chemical rate limited (Laidler, 1987). a. Bimolecular Reactions. In a successful bimolecular collision, part of the kinetic energy of the fast-moving reactant molecules is used to provide the energy of activation and thus to produce the high-energy molecular arrangement of the transition state. TST applied to the reaction, A + B < = = > AB? = => R + S (27)

in which AB? is the transition state structure, leads to the following expression,
? ?E k ? 1.25 x1013T 2 ? S R exp ? a RT ? ? ? ? ?

(28)

the units are cm3/mol-sec. TST shows a T2 dependence on temperature, and the change in entropy leading to the transition state is needed. There is now quantum chemistry software available that make it possible to estimate the properties of the transition state, this will be discussed more fully later. Unfortunately, properties of the transition state cannot yet be tested experimentally, thus the uncertainty in the calculations for the transition state would not be well known. The best approach is to be most familiar with the particular quantum chemistry package that is to be used and its limitations in general. b. Unimolecular Reactions In unimolecular reactions, the necessary energy for the reaction may accumulate in the molecule as the result of intermolecular collisions, photon activation, or as the result of unimolecular chemical activation. Once energy is imparted to the molecule, it is rapidly distributed amongst its vibrational and rotational energy levels with the energized molecule taking many configurations. If one of these configurations corresponds to the localization of enough energy along the reaction coordinate, then the reaction occurs.

The application of TST theory to the process below, A <= = > A ? <= = > P leads to the following expression ( the units are sec-1 )
? ?E k ? 4.60 x1010 T exp ? S R exp ? a RT ? ? ? ? ?

(29)

(30)

TST predicts a first order temperature dependence for the rate coefficient. As with bimolecular reactions, the entropy change leading to the transition state will be required, and quantum chemistry methods may be used for this. c. Lindemanns Approach to Unimolecular Reactions No discussion on chemical kinetic theory would be complete without Lindemanns theory of unimolecular reactions which attempts to explain the pressure dependence of unimolecular reactions.The overall unimolecular reaction is given below,
k uni

A = => B

( 31)

At a given temperature, and for high pressure conditions, the rate of decomposition of A is first order in its concentration, but a low enough pressures, the rate becomes pressure dependent, i.e., the process is energy transfer limited. The dependence of kuni on pressure is shown in Figure 1. In the mechanism developed by Lindemann (Laidler, 1987), the decomposition of reactant A occurs according to the following two step scheme,
k1

A + M < = = > A* + M
k-1 k2

( 32)

A* = = > P

(33)

In reaction (32), molecules of A are energized by collision with a second body M. Reaction (33) describes the process by which the energized molecules of A* turns into the final product. The results of this approach will be given below without a proof, Laidler (1987), Garginer (1972) present a full discussion of Lindemanns Theory.

10

In the high pressure limit the unimolecular rate coefficient takes the form,
? k uni ?

? ?M ?? ? ?

lim kuni ? k 2

k1 k? 1

(34)

whereas at low pressure,

o k uni ? lim k uni ? k1?M ?

(35)

?M ?? ?0?

The high pressure limit does not show a dependence on pressure, in the other hand, the low pressure coefficient is dependent on pressure through the term [M], as it is found experimentally. Estimates of the high pressure limit rate ? coefficient k uni can be made using TST, quantum chemistry can be used to estimate the properties of the transition state. In equation (35), in order to calculate the low-pressure coefficient, k1 is expressed as follows, k1 ? Z A*M ? f ?E0 ?? ? (36)

ZA*M is the molar collisional frequency between energized A* molecules and M (see Equation (25) ), f(E o) is the fraction of molecules with energies higher than Eo and can be activated according to reaction (32), this term may be given in terms of the Boltzmann distribution function P(E) or f ?Eo ? ?
?

Eo

?P?E ?dE

(37)

? is a collisional efficiency that accounts for the fact that not every collision between an activated A* molecule and M results in deactivation of A* back to A. For thermodynamic conditions where unimolecular reactions fall in a regime that is between high and low pressure or the fall-off regime, software is available that can make estimates of the coefficient based on the constants given in equations (34) and (35): Kee et al. (1993), Stewart et al (1989). The reader is referred to these references for more details.

11

3. Quantum-Rice-Ramsperger-Kassel (QRRK) Treatment of Energy Transfer Limited Reactions. Only a brief introduction will be given here to the QRRK treatment of unimolecular and bimolecular reactions. References will be provided for the reader to become more acquainted with this theory as well as software available to carry out the computations needed under the theory. The Lindemann Theory deviates somehow form the experimentally determined behavior of unimolecular reaction (see Figure 1). The reason for this can be explained by discussing what is presented illustrated in Figure 2. The molecule A is activated to A* but in the QRRK treatment, the rate coefficient krxn(E) depends on the excess energy of the activated molecule over the ground state. As the figure shows, QRRK treats the molecular energy as being quantized. A full discussion of this problem can be found in Westmoreland et al. (1986). In a similar manner, for bimolecular reactions (Westmoreland,et al., 1986), the process is depicted in Figure 3. As with unimolecular reactions, the reaction leads to an activated molecule A*. The fate of this molecule depends on its excess energy. The rate coefficient k2(E) for the decomposition to products P + P depends on excess energy over the ground state. The energy is considered as being quantized. More complex schemes involving izomerization of the activated molecule can be found in Kazakov et al. (1994). Software for the mathematical treatment of chemically activated reactions can be found, see for example Dean, Bozzelli, and Ritter (1991) for an introduction to the CHEMACT program and Dean and Westmoreland (1987) for additional information. The reader is referred to these references for a more thorough discussion of chemical activation. 4. Computer Software for Chemical Reactor Engineering. (a) Chemical Kinetics. References were given above for programs that are available to estimate chemically activated reaction. However, very often no information is available for the activation energy of chemical reactions, or properties of the transition state to be able to estimate rate coefficients according to TST. Computer software is now available that can be used to make estimates of such parameters. The software uses the methods of quantum chemistry, a discussion of these methods is beyond the scope of this manuscript, and the reader is referred to Gargurevich

12

(1997), Pople (1970), and Murrell (1972) for a thorough discussion of the theory upon which these programs are based. The software that the author is most familiar with is the Molecular Orbital Package (MOPAC) developed through the Quantum Chemistry Program Exchange at Indiana University. MOPAC is based on semi-empirical methods that use experimental data to arrive at solutions of the equations derived from quantum chemistry. Other properties can be calculated using MOPAC such as the heat of formation of chemical species if needed. Computer software is also available that use ab initio methods, i.e., the equations derived from quantum chemistry are solved using strict theoretical calculations. These require more computer memory and speed than the semi-empirical methods. A list of computer software can be found in www.chemistrysoftware.com/software_guide/modeling_index.htm. As it was stated above, a major problem in studying reactions by any current theoretical models is the lack of experimental data for properties of the transition state. Calculations of these properties then have not been tested, and the performance of the method used for such calculations is safer, the better the performance of the method in question in all areas where it can be tested. (b) Chemical Reactor Design. The reactor simulator that the author is most familiar with is the CHEMKIN package developed by Sandia National Laboratories (R. J. Kee, F. M. Rupley, and J. A. Miller, Sandia National Laboratories Report, SAND89-8003, 1989). CHEMKIN uses different modules in order to carry out the necessary simulation of plug flow reactors, or combustion phenomena for example. These modules are as follows: (1) chemical species and their thermodynamic properties, (2) the elementary reactions composing the complex mechanism, and the reaction rate constants for each reaction, (3) transport properties of each chemical species (this is necessary for the treatment of combustion phenomena). The modules interface with each other to numerically solve the problem at hand. This software was used extensively by the author to simulate combustion phenomena (Gargurevich, 1997). There are other process simulators available that also can be used for the simulation of chemical reactors. The most commonly used are software from Aspen Technology Inc.( Cambridge, Mass.), Simulation Sciences Inc. (Brea,

13

Calif.), Hyprotech ( Calgary, Alberta) , and Chemstations Inc. (Houston, Texas) (Simulators seek a broader community of users, Chemical & Engineering News, March 27, 1995). The author is not familiar with the use of these simulators in the design of chemical reactors except for PRO II 5.5 (Simulations Science) for other applications. PRO II has a plug flow reactor unit operation plus others which allow the user to specify the reactions and the rate coefficients for each reaction. LEAST COMPLEX CHEMICAL RATE LIMITED (1) Simple Kinetic Theory of bimolecular reactions: reaction rate coefficient equals the collision frequency. (2) Modified Kinetic Theory: reaction rate coefficient equals, collision frequency x activation energy factor x steric factor (3) a. Transition State Theory (TST): unimolecular/bimolecular reactions. Rate coefficient includes an activation energy factor, and an entropy factor to account for steric effect. b. Lindemann theory of unimolecular reactions.

ENERGY TRANSFER LIMITED. (4) Quantum mechanical treatment of thermal activation, and unimolecular/bimolecular chemical activation. Rate coefficients depend on the internal energy of molecular species.

14

MOST COMPLEX

Table 1. Theories for estimating reaction rate coefficients.

Reaction Type I.Unimolecular Reactions, k= A exp(-Ea/RT), s -1 a. Simple Fission Reactions CH4 < = = > CH3 +H Recommendation b. Complex Fission Reactions b1. Radical Fissions C2H5 < = => C2H4 +H Recommendation CH3CO <= = > CH3 + CO Recommendation b2. Molecular Fissions Three and Four Center Eliminations C2H5Cl < = = > C2H4 + HCl Recommendation b3. Isomerization Reactions b3.1 Cis- Trans Isomerization CHCl=CHCl cis < = = > trans Recommendation b3.2 Atom Migrations 1-C3H7 <= = > 2-C3H7 Recommendation b3.3 Cyclizations & Decyclizations Recommendation Log10 A Ea (Kcal/mole)

? Hr (Kcal/mole) @298 K

16.0 15.5

105 ? Hr

105

13.6 13.0 12.4 13.0

40.9 ? Hr 16.7 ? Hr +5

38.7 14.2

13.5 13.5

56.6 1/3( sum of bond energies broken) or ? Hr + 25

20.2

13.8 13.0 12.4 13.0 15.2 15.0

63.0 Pi-bond energy + 4.0 34.0 35 for 1,2 shift 25 for 1,3 shift 65.0 60.0

0.0

0.0

-27.6

15

Table 2. High Pressure Rate Coefficients for Analogous Reactions (Senkan, 1992
Reaction Type II Bimolecular Reactions K= AT 2exp(Ea/RT) cm 3/mol-sec a. Atom Metathesis Reactions H + CH4 < = = > CH3 + H2 Recommendation CH3 + C2H6 < = = > CH4 + C2H5 Recommendation Reaction A + BC , = = > AB + C Group or Atom H Cl O OH NH2 HO2 CHO CH3 C2H5 b. Radical-radical Metathesis C2H3 + H < = = > C2H2 +H2 Recommendation c. Molecule -Molecule Metathesis C2H4 + C2H4 < = = > C2H3 + C2H5 Recommendation: Use microscopic reversibility. III.Complex Bimolecular Reactions k= Aexp(Ea/RT) cm 3/mol-sec CH3 + O2 <= = > CH2O + OH Recommendation: None. Log10 A Ea (Kcal/mole) ? Hr (Kcal/mole) @298 K

8.85 8.50 6.55 7.00

11.6 FA x FC see below 10.8 FA x FC see below

0.0

-4.7

F 3.00 0.57 2.15 1.30 1.30 1.70 1.55 3.50 2.85 8.35 8.00 9.15 0.0 0.0 72.0 -66.3

70.0

13.7

34.6

-53.0

Table 2. High Pressure Rate Coefficients for Analogous Reactions (Senkan, 1992).

16

References Arakawa, S. T. et al., Increase Productivity with Novel Reactor design, Hydro. Proc., March 1998, p. 93. Baulch, D. et al., Evaluated Kinetic Data for Combustion Modeling, J. Phys. Chem. Ref. Data, Vol 21, No.3, 1992. Chang, R., Chemistry, Chap. 13, MacGraw-Hill, Inc., 1994. Dean, A. M., and Westmoreland, P. R., Bimolecular QRRK Analysis of Methyl Radical Reactions, International Journal of Chemical Kinetics, 19, 3, pp. 207228, 1987. Dean, A. M., Bozzelli, J. W., and Ritter, E. R., CHEMACT: a Computer Code to estimate Rate Gas Constants for Chemically Activated Reactions, Comb. Sci. Tech., 80, 1-3, pp. 63-85, 1991. DeMore, W. B. et al., Chemical Kinetics and Photochemical Data for Use in Stratospheric Modeling, Evaluation No. 9, JPL Publication 90-1, Jan. 1990. Dryer, F. L., The Phenomenology of Modeling Combustion Chemistry, in Fossil Fuel Combustion: A Source Book, pp. 121-213, John Wiley & Sons, New York, 1991. Gardiner, W. C., Rates and Mechanisms of Chemical Reactions, W. A. Benjamin Inc., 1972. Gargurevich, I. A., Combustion Models and the Prediction of Policyclicic Aromatic Hydrocarbons (PAH) in Laminar Flames of Simple Hydrocarbons: Methane, Methyll Chloride, and 1,1 Dichloroethane , Ph.D. Dissertation, UCLA Chemical Engineering Department, 1997. Kazakov, A., Wang, H., and Frenklach, M. , J. Phys. Chem., 98, p. 10598, 1994. Kee, R. J. et al., Chemkin-II: A Fortran Chemical Kinetics Package for the Analysis of Gas Phase Chemical Kinetics, Sandia National Laboratory Report, SAND 89-8009B, UC-706, 1993. Laidler, K. J., Chemical Kinetics, Harper & Row, New York, 1987. Miller, J. A., and Bowman, C. T., Prog. Energy Combust. Sci., Vol 15, pp.287338, 1989.

17

Miller, J. A., and Melius, C. F., Combust. And Flame, 91, pp.21-39, 1992. Murrell, J. N., and Harget, A. J., Semi-Empirical Self-Consistent-Field Molecular Orbital Theory of Molecules, Wiley-Interscience, New York, 1972. National Institute of Standards (NIST) Reference Database 17, NIST Chemical Kinetics Database 60, 1994. Pople, J. A., and Beveridge, D. L., Approximate Molecular Orbital Theory, McGraw Hill, New York, 1970. Qun, M., and Senkan, S. M., Combust. Sci. Tech., vol. 101, pp. 103-134, 1994. Seinfield, J. H., Science, 243, p. 745, 1989. Senkan, S. M., Detailed Chemical Kinetic Modeling: Chemical Reaction Engineering of the Future, in Advances in Chemical Engineering, vol. 18, pp. 95115, Academic Press, 1992. Senkan ,S. M., Survey of Rate Constants in the C/H/Cl/O System, in Combustion Chemistry, 2 nd edition, W. C. Gardiner Jr. editor, 1993. Stewart, P. H., Larson. C. W., and Golden, D., Combust. And Flame, 75, p.25, 1989. Wang, H., Detailed Chemical Kinetic Modeling of Soot Particle Formation in Laminar Premixed Hydrocarbon Flames, Ph.D. Thesis, Pennsylvania State University, August 1992. Warnatz, J., Rate Coefficients in the C/H/O System, in Combustion Chemistry, edited by W. C. Gardiner, Spring -Verlag, New York, 1984. Westbrook, C. K., and Dryer, F. L., Prog. Energy Combust. Sci., 10, pp. 1-37, 1984. Westmoreland, P. R., Howard, J. B., and Longwell, J. P., AICHE Journal, 32, p. 1971, 1986. Worstell, J. H., Dont Act Like a Novice About Reaction Engineering, Chem. Eng. Prog., March 2001, p68.

18

Lindemanns kinf

kuni

Experimental
k0 uni

Absolute Pressure @ Const. Temperature

Figure 1. Pressure Dependence of k uni

19

A* n

k ?E ? rxn

kexc

Product(s)

Figure 2. Energy Dependence of kuni (Westmoreland (1986).

20

k? 1 ?E ?

prod

?E ?

k1 R + R

A*

P + P

Reaction Coordinate

Figure 3. Bimolecular Chemical Activation (Westmoreland (1986).

21